Non-dispersive infrared (NDIR) detectors may typically comprise an IR source, a sample chamber (containing the gas sample), a sample detector, and a reference detector. The detectors may comprise optical bandpass filters depending on the target gas(s). The sample detector is used to detect the target gas and the reference detector is used to ignore the target gas and any known interferrants. The reference detector provides a base point or zero while the sample detector provides the signal with the differential providing the actual span value of the instrument. This sample/reference approach compensates for the changes that can occur in the detector sensitivity or source. For example, the source intensity can change due to contamination causing a zero drift.
It is a common safety practice to use two detectors with a means of selecting different wavelength bands of the source light. For example, the reference signal can be used in conjunction with the sample signal to determine any drop in the intensity of the radiation output, any loss of intensity due to fouling of the detector (e.g., a fogged or dirty window, etc.), or any substances in the light path that may affect the intensity of the radiation (e.g., dust, water vapor, etc.). The reference detector can also be used to ensure that radiation is being received. If the reference detector does not have a signal, then an indication that the radiation is not present may be generated. This may help ensure that the system is operating. In comparison, a zero response in a prior system may simply be interpreted as a lack of the presence of a target gas when in fact the light source is not working. The reference signal can be used to compensate the detected signal from the sample detector to produce a response with an improved accuracy.